Computing is moving to the edge of the cloud
ICT executives and managers need to consider a dramatic shift in computing architecture that promises to disrupt cloud computing.
In the last decade cloud computing has grown dramatically. Traditional corporate computing leaders like IBM and Microsoft are struggling to catch up to an upstart, Amazon Web Services. The cloud has given birth to new markets and buzzwords like hyper converged infrastructure, software-defined networking, and containers.
Notwithstanding its successes, a growing number of experts are predicting that the cloud will now itself be transformed by what is called "edge computing."
Edge computing envisions that “smart” connected objects, such as the billions of devices connected to the Internet of Things (IoT), will perform more processing and storage, replacing the centralized architecture of today’s cloud.
Consider future millions of self-driving cars, each with 200-plus processors generating 25GB of data every hour. Each will need to get computational results immediately, so the vehicle knows when to turn or brake. Now add to self-driving cars billions of other intelligent devices: drones, medical equipment, video cameras, manufacturing robots etc. There will never be enough network bandwidth and speed to handle the traffic they generate. Add their enormous need for data storage—and remember that storage is more prone to failure than processing- and you have a problem that the cloud alone cannot solve. Or consider a wearable device like Fitbit that generates a lot of data but has limited storage and battery life because its flash memory chips consume a lot of space and power. How will all this data efficiently flow through the network?
The problem is the cloud is usually far away from where data originates and processing results like real-time analytics are needed. And the cloud is prone to latency. Therefore, devices at the edge of the network near the source of the data will have to do more processing themselves, sending to the cloud only the information that most needs to be analyzed and shared with other devices.
The answer will be to distribute computing power and the related cost between the edge and the cloud. Edge computing platforms will rationalize what data to transmit, store and analyze locally, such as real-time analytics that impact the bottom line, prevent a disaster or fraud or fight data breaches in real time. Everything else it will send to the cloud.
Case in point of a product that is seizing the edge computing opportunity: MachineShop, a fast-growing edge computing platform, which automates and orchestrates collection, analysis and integration of operational data locally and then provides a simple, seamless link to virtually any cloud application or platform. Locally configurable systems like MachineShop are disrupting cloud architecture and are certainly a harbinger of things to come.
Business leaders should expect to hear from their ICT organization about the need to invest in R&D and prototyping of new edge computing systems. It will be an investment worth making.
Leave a comment
Your e-mail address will not be published. Required fields are marked *